A new global analysis has linked long-term exposure to three widespread air pollutants with millions of newly diagnosed cancer cases, offering one of the most extensive assessments yet of how polluted air may shape the worldwide cancer burden. The study, published in Nature Health, examined 109 million cancer cases recorded across 952 locations between 2000 and 2020. Researchers focused on fine particulate matter, or PM₂.₅, ozone and nitrogen dioxide—pollutants produced by traffic, industry, power generation, household combustion and chemical reactions in the atmosphere. Their findings suggest that the cancer risks associated with air pollution extend far beyond the respiratory system and may affect populations on every continent.
The strongest association was observed for PM₂.₅, the microscopic particles measuring 2.5 micrometres or less in diameter. Because these particles are so small, they can penetrate deep into the lungs and, in some cases, cross into the bloodstream. Their chemical composition varies according to their source, but PM₂.₅ may contain metals, soot, organic compounds and other substances capable of triggering oxidative stress and chronic inflammation. Over time, these biological disturbances can damage DNA, alter immune responses and interfere with normal cellular repair. The new analysis found that every additional 10 micrograms of PM₂.₅ per cubic metre of air was associated with a 16.0 percent increase in the risk of all cancers combined.
The study also identified a J-shaped exposure–response relationship for PM₂.₅. In epidemiology, this pattern indicates that the risk does not rise in a simple straight line across all exposure levels. At lower concentrations, changes in risk may be relatively modest or difficult to distinguish, while at higher concentrations the curve turns upward more sharply. A J-shaped association can reflect biological thresholds, differences in population susceptibility or the effects of particularly intense pollution exposure. It may also emerge when background risks and other environmental factors vary across locations. The researchers used statistical models designed to capture this type of nonlinear pattern rather than assuming that every incremental increase in pollution carries exactly the same effect.
Ozone, commonly known as O₃, showed a different pattern. Unlike the protective ozone layer high in the atmosphere, ground-level ozone is a harmful pollutant formed when nitrogen oxides and volatile organic compounds react in sunlight. It is often more severe during hot, sunny conditions and can travel across administrative boundaries, making it difficult for individual cities to control. Ozone irritates the airways and can promote inflammation throughout the respiratory system. According to the study, each 10 micrograms per cubic metre increase in long-term ozone exposure was associated with a 4.23 percent rise in the risk of all cancers. The relationship was described as near-linear, meaning that the estimated risk increased more steadily as exposure rose.
Nitrogen dioxide, or NO₂, was associated with an 11.7 percent increase in all-cancer risk for every 10 micrograms per cubic metre increase in long-term exposure. The gas is generated primarily by fuel combustion, especially from vehicles, power plants and industrial activity. It can damage airway tissues directly and also contributes to the formation of ozone and secondary particles. Nitrogen dioxide is therefore both a pollutant in its own right and part of a wider atmospheric chemical system. Exposure is frequently highest near busy roads and in densely populated urban areas, where traffic emissions can create sharp differences in air quality from one neighbourhood to another.
When the researchers translated these relative risks into population-level estimates, the scale of the findings became considerably larger. PM₂.₅ exposure was estimated to be associated with approximately 8.82 million incident cancer cases worldwide during the study period. Ozone was linked to about 2.59 million cases, while nitrogen dioxide was associated with approximately 6.67 million cases. These figures represent attributable burdens calculated from population exposure and estimated risk relationships; they do not mean that every individual case can be traced to a single pollutant. Cancer is a multifactorial disease influenced by age, genetics, smoking, alcohol use, infections, diet, occupational hazards and access to medical care. Nevertheless, even a modest increase in risk can produce a substantial number of cases when it affects billions of people.
The researchers used distributed lag non-linear models, a method that can evaluate both delayed effects and nonlinear exposure patterns. Cancer often develops over years or decades, so the consequences of an exposure may not appear immediately after pollution levels rise. Distributed lag models allow investigators to examine how risk may accumulate across time rather than assigning all effects to the moment of exposure. The approach can also account for changing exposure levels, differences between locations and potential variations in the time required for environmental damage to contribute to disease. By combining these models with global cancer and pollution data, the study aimed to estimate not only relative risks but also the number of cases potentially linked to ambient air pollution.
One of the study’s most striking observations was the consistently higher attributable risk and burden among women for most malignancies. The analysis does not establish a single explanation for this difference, and the causes may vary by region and cancer type. Women and men can experience different occupational and household exposures, including pollution from cooking fuels, heating systems and poorly ventilated indoor environments. Biological differences in hormone regulation, immune function, body composition and pollutant metabolism could also influence susceptibility. In addition, patterns of healthcare access, diagnosis and cancer registration may affect the data. The finding highlights the importance of examining air pollution through a sex-specific lens rather than assuming that the same exposure produces identical outcomes in every population.
The results arrive as cities worldwide confront overlapping challenges from traffic emissions, industrial pollution, wildfires and climate-driven heat extremes. Fine particles can travel long distances, while hot weather can intensify ozone formation and increase the frequency of stagnant-air episodes. Reducing fossil-fuel combustion, improving public transport, tightening industrial standards, expanding clean household energy and monitoring pollution at neighbourhood scale could therefore deliver benefits beyond cardiovascular and respiratory health. The authors describe their estimates as an impetus for stronger public-health policies aimed at reducing the global cancer burden. Although the study is observational and its estimates depend on the quality of exposure data, cancer records and statistical assumptions, its vast geographic scope underscores a central message: cleaner air may be one of the most consequential cancer-prevention measures available to governments and communities.
Subject of Research: Global cancer burden associated with long-term exposure to ambient air pollution
Article Title: Global cancer burden associated with long-term exposure to ambient air pollution
Article References: Zhang, G., Li, Y., Li, A. et al. Global cancer burden associated with long-term exposure to ambient air pollution. Nat. Health (2026). https://doi.org/10.1038/s44360-026-00180-4
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s44360-026-00180-4
Keywords: Air pollution, PM₂.₅, ozone, nitrogen dioxide, cancer, public health, environmental epidemiology, global health, attributable burden
